Significance by Sight

The book aims to re-establish graphical analysis and simple correlation checks as a necessary first step, a prerequisite, in empirical economic and social science research. Its value lies in challenging the widespread practice of moving directly to complex econometric estimation without first verifying whether the core variables exhibit any meaningful and obvious empirical relationship.

For its intended readership, applied economists, econometricians, and policy analysts, the book offers a unifying diagnostic principle illustrated across several major literatures, with the promise of improving inference quality and avoiding misspecification. The author argues that any economic research questions’ causal nature can be seen clearly in a scatterplot. However, if there are no discernible associations graphically and yet there are statistically significant findings reported in the paper, it raises some doubts about parts of the paper’s findings.

Using this backdrop, the book picks a few fundamental areas economics, and within each area, the book chooses a few seminal papers’ questions and represent their research questions’ association graphically (using scatterplots), either to support or question their hypotheses. The intended readers, after looking at a bunch of fundamental and seminal papers through the lens of simple scatterplots, would have a cleaner idea on how to critique and appreciate those papers. This is indeed a fundamental skill that every student, especially those who use or intend to use data extensively should have. Yet, there are only a few guidelines out there. Along with this, reading a set of seminal empirical papers, seemingly from unrelated areas of economics using one simple approach, a correlational and graphical lens, is a treat for students, and economics scholars.

Key Features:

• Includes extensive statistical and graphical analysis of eight fundamental research questions in macroeconomics, international finance, monetary economics, and microeconomics
• Focuses on graphical analysis as a prerequisite for applied and econometric studies
• It’s primarily about seeing plots and correlations of time series data first then deciding whether there is a value in further analyzing the data, or not